Analysis
Three companies attacking three different layers of the same underlying problem have each reached nine- or ten-figure valuations within months of launching. Groundcover raised a $100 million Series C led by One Peak at a $500 million valuation, betting that AI agents -- not human engineers -- are now observability tooling's primary users, putting it in direct competition with incumbent Datadog. Glow emerged from stealth with $180 million raised across three rounds in roughly a year, reaching a $1.2 billion valuation to rebuild endpoint security specifically for AI-native threats. Onyx Security raised $113 million just four months after its own stealth launch, reaching a roughly $640 million valuation for a platform that monitors and governs what autonomous agents actually do once they're already inside enterprise software.
What's notable is that none of these three is consolidating the others' territory. Groundcover is rebuilding observability around machine consumers instead of human ones; Glow is protecting the endpoint itself against AI-era threats; Onyx is governing agent behavior after an agent already has access. That the category is fragmenting into specialized layers this early, rather than one platform absorbing the rest, tells you investors see enough distinct surface area to fund multiple winners rather than a single consolidator.
“What's notable is that none of these three is consolidating the others' territory.”
The speed is the real signal. Two years ago, agent oversight barely existed as a line item in an enterprise security or infrastructure budget. Now three separate companies have each cleared a nine-figure valuation within roughly a year or less of launch -- a compression that mirrors what happened across most AI-native software categories in 2026, but is notable here because it means enterprise buyers are already carving out dedicated spend for this, not just experimenting with a feature inside an existing tool's renewal.
For founders and investors in adjacent categories, the read-through is that "agent governance" isn't a feature checkbox anymore -- it's graduating into its own procurement category with its own budget owner, the way "cloud security posture management" did a decade ago. What to watch: whether incumbents like Datadog, CrowdStrike or Wiz respond by building agent-native interfaces themselves, or whether they let this generation of specialists own the category the way legacy SIEM vendors ceded ground to cloud-native security startups the last cycle.